www.ijecs.in International Journal Of Engineering And Computer Science ISSN:2319-7242

advertisement
www.ijecs.in
International Journal Of Engineering And Computer Science ISSN:2319-7242
Volume 3 Issue 7 July, 2014 Page No. 7239-7241
Role of Data Mining in the Manufacturing and
Service Application
Rajesh Kumar
bmbpnt@gmail.com
Abstract: In the today scenarios, the manufacturing and service industries are focusing on the highly effective tools which are mainly
concentrated over the data mining. Several organisations and researchers have accomplished their efforts in the reviewing of the data
mining tools and the surveying the data miners. In this present paper, the notion of machine learning and the data mining in the various
manufacturing and service application is introduced. On the result, the idea of building the decision making systems are more clear as the
algorithms of the machine learning helps in extracting the knowledge from the databases which are in the diverse form. As in the case
where the operational engineering data is based, we can determine the optimal control parameters, faults detection in the equipment. So, the
framework is designed which uses constructs such as decision maps, decision tables, etc. for the decision making system that helps in
applying and organizing knowledge in the manufacturing and service application.
Keywords: Decision making, data mining, knowledge structuring, industrial application, modelling.
1. Introduction
There are rise in the amount of enterprise data effecting and
presenting many issues which deals with extracting, storing
organizing and using the knowledge which is generated from
the data sets. As we are aware of that the data is often
distributed and heterogeneous so the understanding of the
meaning and structure of the stored information is not very
easy. There is diverse capacity of complexness in the reality
but in spite of this, the decision making is recognized as by the
help of decision making tool, the set of previously composed
data is processed. The decision would not be relevant if the
decision making tool is not properly versed in the current data.
There are two types of decision making i.e. stand alone
decision making and distributed decision making.
The decision making system faced many boundaries which has
now eliminated as there are many progresses in the
development of networking technology and data analysis tools.
The task of extraction of knowledge from the heterogeneous
and distributed sources of data is played by the field of data
mining. In the distributed data mining the different scenarios
are probable to take place. For instance,
 At the single or multiple sites, the extraction of
knowledge may be distributed by the intermediate and
final decision.
 The cycles may form during the flow of data in spite of
being linear.
2. Descriptive Data Mining Application
Figure 1: Stand alone decision making
Figure 2: Distributed decision making
In this paper, the following will explained the descriptive
examples of the service and manufacturing application where
the data mining have been powerfully accomplish the role.
 Manufacturing of semi conductor
Nowadays, the main area of the manufacturers of the
semi conductor is in the improvement of the wafers
quality. At the different various plants which produce
the wafer, the manufacturing environment, may be
alike; some wafers attain perfect quality and some fail
in it. For the desired quality, the data mining is
performed to determining the parameters ranges.
 Chemical process and biotechnology industry
In the field of the bio and chemical industry, the
processes uses the data mining and could be well
equipped in data collection systems as usually they
cannot be effectively modeled by the classical tools.
Rajesh Kaumar, IJECS Volume 3 Issue 7 July, 2014 Page No.7239-7241
Page 7239
 Assembly of electronic
Similar to the semi conductor manufacturing industry,
the electronic product assembly has facing the
problems which deals with the quality of the printed
circuit boards. Their components fail the test of
quality for the unidentified reasons. So the prevention
of the failure is of great interest to the manufacturers,
so far this issue the data mining has been applied to
identify the process parameter’s condition by
minimizing the faulty products production. There are
many other application where such challenges are
faced by the manufacture of different industries and
areas of different fields. For instance, DNA
manufacturing,
energy
production,
medical
applications, pharmaceutical applications.
3. Extraction Of Knowledge
On the increasing of the awareness in the era of the data mining
has taking the steps to the development of different algorithms
which extract rules, knowledge, feature from the large volume
of sources. By the machine learning algorithm, these
developments in the era of data mining have not been matched
the progress in decision making which are based on knowledge
extraction. Somehow, the most data mining outputs shown have
been developed by the non-engineering community and they
mainly have the prime objective in the extraction of knowledge
in spite of the decision making.
Learning algorithms are categorized as follows which mainly
become the interest for the industry:
 Decision rule algorithm
 Induction logic program algorithm
 Rough set algorithm
 Decision tree algorithm
3.1 Evaluation of Features
In data mining, this paper fills the gap in the decision making
which has been created by the unique progress. The framework
for decision making and the algorithm presented for the
knowledge extraction use by algorithm of data mining.
Numbers of application areas are seen like quality engine,
equipment diagnostics, medical diagnosis, planning and
forecasting look forward to decision making tools.
The following matrices are used for the data sets and features
evaluation:
 Classification Quality
 Gini index
 Upper and lower approximation measures
 Entropy measure
The process of extraction of knowledge leads to the decision
rules (IF...THEN rules) which are formulated from the data
sets. In spite of making decisions directly with the decision
rule, we have also the decision making approaches which are
more transparent and structured to the user.
As the amount of data sets increase , the rules which are
desired by the machine learning algorithm for the decision
making are also becoming complex. To minimize this strong
complexity, the method and approach are built upon which are
transparent to the customers by using the new thoughts of
decision map, decision atlas and decision table. These are
helpful in the management of knowledge and structuring of the
different information.

Decision maps, tables and atlases
A decision map is defined as the decision process
model and the collection of decision tables. At any
step, the decision map, the actual decision which can
be performed by using the content of one or more
decision tables. The collection of the knowledge
needed to make the decision in a particular area is
known as the decision table. The decision maps are
merged in an atlas and these multiple atlases create a
library. The decision may be accomplish in the
various manner based on the decision tables.

Process modelling and decision making
In the field of data driven decision making, the idea of
a process and a modelling methodology plays an
important role. For modelling a process, the
methodology which is selected may rely on many
factors which range from culture of organisation and
training to the availability and correctness of the
method.

Decision making algorithms
The main objective is to generate (100-α) percent accurate
decision for (100-β) percent conditions with identified
decision. The notion of the orthogonal algorithms P
(primary) and C (confirmation) algorithm is presented in
this paper which allows α meeting probably to 0 and β to
remains tiny. The decision using the structured knowledge
which is derived from the different learning algorithm is
generated from the P-algorithm and the same decision
using an orthogonal concept like classical algorithm, a
distance metric based algorithm is derived Calgorithm.
4. Decision Making
Rajesh Kaumar, IJECS Volume 3 Issue 7 July, 2014 Page No.7239-7241
Figure 3: The logic of decision making
Page 7240
Both of the P and C- algorithm are observed as the parallel
system component so contributing to the whole system
accuracy.
References
[1]
5. Data Transformation
The decision making algorithm quality mainly relies on the
knowledge extracted from data sets. As these data sets are
mined in their raw form so they are needed to be transformed.
The aspect of the data transformation is used in manufacturing
and service application for transparent, to improve usability
and for the decision making accuracy.
6. Conclusion
The present study has concluded that the increase in the amount
of data in the manufacturing and service industries has been
facing some issues which are needed to discover the tool that
can create analysis of data and knowledge. Somewhere, the
significant progress has taken place in the growth of algorithm
for extracting knowledge. The proposed approach for the
decision making is likely to be used for the application in many
areas of the manufacturing and service industries.
[2]
[3]
[4]
[5]
Durán, O., Rodriguez, N.
& Consalter, L.A.
(2010).Collaborative article swarm optimization with a
data mining technique for manufacturing cell design.
Expert Systems with Applications 37, 1563-1567.
Chien, C., Wang, W. & Cheng, J. (2007).Data mining for
yield enhancement in semiconductor manufacturing and
an empirical study. Expert Systems with Applications
33, 192–198.
De Jong, K.A.,Evolutionary Computation: A Unified
Approach, 2006 (MIT Press:Cambridge, MA).
Hall, M., E. Frank, G. Holmes, B. Pfahringer, P.
Reutemann, and I. H. Witten. The weka data mining
software: An update.SIGKDD Explorations, 11, 2009.
Mierswa, I., M. Wurst, R. Klinkenberg, M. Scholz, and T.
Euler. Yale: Rapid prototyping for complex data mining
tasks.In Proceedings of the 12th ACM SIGKDD
International Conference on Knowledge Discovery and
Data Mining (KDD-06), 2006.
Rajesh Kaumar, IJECS Volume 3 Issue 7 July, 2014 Page No.7239-7241
Page 7241
Download